---
title: "Coronavirus in Denmark"
author: "Tobias Søndergaard"
output: 
  flexdashboard::flex_dashboard:
    orientation: rows
    # social: ["facebook", "twitter", "linkedin"]
    source_code: embed
    vertical_layout: fill
    
knit: (function(input_file, encoding) {
  out_dir <- 'docs';
  rmarkdown::render(input_file,
 encoding=encoding,
 output_file=file.path(dirname(input_file), out_dir, 'index.html'))})
    
---

```{r setup, include=FALSE}
#------------------ Packages ------------------
library(flexdashboard)
# install.packages("devtools")
#devtools::install_github("RamiKrispin/coronavirus", force = TRUE)
library(coronavirus)
data(coronavirus)
# View(coronavirus)
# max(coronavirus$date)
`%>%` <- magrittr::`%>%`
#------------------ Parameters ------------------
# Set colors
# https://www.w3.org/TR/css-color-3/#svg-color
confirmed_color <- "purple"
active_color <- "#1f77b4"
recovered_color <- "forestgreen"
death_color <- "red"
#------------------ Data ------------------
df <- coronavirus %>%
  # dplyr::filter(date == max(date)) %>%
  dplyr::filter(country == "Denmark") %>%
  dplyr::group_by(country, type) %>%
  dplyr::summarise(total = sum(cases)) %>%
  tidyr::pivot_wider(
    names_from = type,
    values_from = total
  ) %>%
  # dplyr::mutate(unrecovered = confirmed - ifelse(is.na(recovered), 0, recovered) - ifelse(is.na(death), 0, death)) %>%
  dplyr::mutate(unrecovered = confirmed - ifelse(is.na(death), 0, death)) %>%
  dplyr::arrange(-confirmed) %>%
  dplyr::ungroup() %>%
  dplyr::mutate(country = dplyr::if_else(country == "United Arab Emirates", "UAE", country)) %>%
  dplyr::mutate(country = dplyr::if_else(country == "Mainland China", "China", country)) %>%
  dplyr::mutate(country = dplyr::if_else(country == "North Macedonia", "N.Macedonia", country)) %>%
  dplyr::mutate(country = trimws(country)) %>%
  dplyr::mutate(country = factor(country, levels = country))
df_daily <- coronavirus %>%
  dplyr::filter(country == "Denmark") %>%
  dplyr::group_by(date, type) %>%
  dplyr::summarise(total = sum(cases, na.rm = TRUE)) %>%
  tidyr::pivot_wider(
    names_from = type,
    values_from = total
  ) %>%
  dplyr::arrange(date) %>%
  dplyr::ungroup() %>%
  #dplyr::mutate(active = confirmed - death - recovered) %>%
  dplyr::mutate(active = confirmed - death) %>%
  dplyr::mutate(
    confirmed_cum = cumsum(confirmed),
    death_cum = cumsum(death),
    recovered_cum = cumsum(recovered),
    active_cum = cumsum(active)
  )
df1 <- coronavirus %>% dplyr::filter(date == max(date))
df_daily <- df_daily[df_daily$date >= "2020-02-27",]
```

Summary
=======================================================================

Row {data-width=400}
-----------------------------------------------------------------------

### confirmed {.value-box}

```{r}
valueBox(
  value = paste(format(sum(df$confirmed), big.mark = ","), " (",
    round(100 * sum(df$recovered, na.rm = TRUE) / sum(df$confirmed), 1), "%)", sep = ""),
  caption = "Total confirmed cases (recovery rate)",
  icon = "fas fa-user-md",
  color = confirmed_color
)
```
















### death {.value-box}

```{r}
valueBox(
  value = paste(format(sum(df$death, na.rm = TRUE), big.mark = ","), " (",
    round(100 * sum(df$death, na.rm = TRUE) / sum(df$confirmed), 1),
    "%)",
    sep = ""
  ),
  caption = "Deaths (death rate)",
  icon = "fas fa-heart-broken",
  color = death_color
)
```


Row
-----------------------------------------------------------------------

### **Daily cumulative cases by type** (Denmark only)
    
```{r}
plotly::plot_ly(data = df_daily) %>%
  plotly::add_trace(
    x = ~date,
    # y = ~active_cum,
    y = ~confirmed_cum,
    type = "scatter",
    mode = "lines+markers",
    # name = "Active",
    name = "Confirmed",
    line = list(color = active_color),
    marker = list(color = active_color)
  ) %>%
  plotly::add_trace(
    x = ~date,
    # y = ~active_cum,
    y = ~recovered_cum,
    type = "scatter",
    mode = "lines+markers",
    # name = "Active",
    name = "Recovered",
    line = list(color = recovered_color),
    marker = list(color = recovered_color)
  ) %>%
  plotly::add_trace(
    x = ~date,
    y = ~death_cum,
    type = "scatter",
    mode = "lines+markers",
    name = "Dead",
    line = list(color = death_color),
    marker = list(color = death_color)
  ) %>%
  plotly::add_annotations(
    x = as.Date("2020-02-27"),
    y = 1,
    text = paste(
        "First case in DK",
        "
",
        "Feb 27th"),
    xref = "x",
    yref = "y",
    arrowhead = 5,
    arrowhead = 3,
    arrowsize = 1,
    showarrow = TRUE,
    ax = 0,
    ay = -90
  ) %>%
  plotly::add_annotations(
    x = as.Date("2020-03-11"),
    y = 800,
    text = paste(
      "Covid-19 declared",
      "
",
      "a pandemic"
    ),
    xref = "x",
    yref = "y",
    arrowhead = 5,
    arrowhead = 3,
    arrowsize = 1,
    showarrow = TRUE,
    ax = 0,
    ay = -90
  ) %>%
  plotly::add_annotations(
    x = as.Date("2020-03-17"),
    y = 1250,
    text = "Denmark is locked down",
    xref = "x",
    yref = "y",
    arrowhead = 5,
    arrowhead = 3,
    arrowsize = 1,
    showarrow = T,
    ax = 0,
    ay = -90
  ) %>%
  plotly::add_annotations(
    x = as.Date("2020-04-01"),
    y = 14,
    text = paste(
      "SSI starts calculating",
      "
",
      "# of recovered"
    ),
    xref = "x",
    yref = "y",
    arrowhead = 5,
    arrowhead = 3,
    arrowsize = 1,
    showarrow = T,
    ax = 0,
    ay = 60
  ) %>%
  plotly::add_annotations(
    x = as.Date("2020-04-14"),
    y = 14,
    text = "Denmark slowly reopens",
    xref = "x",
    yref = "y",
    arrowhead = 5,
    arrowhead = 3,
    arrowsize = 1,
    showarrow = T,
    ax = 0,
    ay = 60
  ) %>%
  plotly::layout(
    title = "",
    yaxis = list(title = "Cumulative number of cases"),
    xaxis = list(title = "Date"),
    legend = list(x = 0.1, y = 0.9),
    hovermode = "compare"
  )
```

Comparison
=======================================================================


Column {data-width=400}
-------------------------------------


### **Daily new cases**
    
```{r}
daily_confirmed <- coronavirus %>%
  dplyr::filter(type == "confirmed") %>%
  dplyr::filter(date >= "2020-02-27") %>%
  dplyr::mutate(country = country) %>%
  dplyr::group_by(date, country) %>%
  dplyr::summarise(total = sum(cases)) %>%
  dplyr::ungroup() %>%
  tidyr::pivot_wider(names_from = country, values_from = total)
#----------------------------------------
# Plotting the data
daily_confirmed %>%
  plotly::plot_ly() %>%
  plotly::add_trace(
    x = ~date,
    y = ~Denmark,
    type = "scatter",
    mode = "lines+markers",
    name = "Denmark"
  ) %>%
  plotly::add_trace(
    x = ~date,
    y = ~Norway,
    type = "scatter",
    mode = "lines+markers",
    name = "Norway"
  ) %>%
  plotly::add_trace(
    x = ~date,
    y = ~Sweden,
    type = "scatter",
    mode = "lines+markers",
    name = "Sweden"
  ) %>%
  plotly::add_trace(
    x = ~date,
    y = ~Finland,
    type = "scatter",
    mode = "lines+markers",
    name = "Finland"
  ) %>%
  plotly::layout(
    title = "",
    legend = list(x = 0.1, y = 0.9),
    yaxis = list(title = "Number of new confirmed cases"),
    xaxis = list(title = "Date"),
    # paper_bgcolor = "black",
    # plot_bgcolor = "black",
    # font = list(color = 'white'),
    hovermode = "compare",
    margin = list(
      # l = 60,
      # r = 40,
      b = 10,
      t = 10,
      pad = 2
    )
  )
```
 
### **Cases distribution by type**

```{r daily_summary}
df_EU <- coronavirus %>%
  # dplyr::filter(date == max(date)) %>%
  dplyr::filter(country == "Denmark" |
    country == "Norway" |
    country == "Finland" |
    country == "Sweden") %>%
  dplyr::group_by(country, type) %>%
  dplyr::summarise(total = sum(cases)) %>%
  tidyr::pivot_wider(
    names_from = type,
    values_from = total
  ) %>%
  # dplyr::mutate(unrecovered = confirmed - ifelse(is.na(recovered), 0, recovered) - ifelse(is.na(death), 0, death)) %>%
  dplyr::mutate(unrecovered = confirmed - ifelse(is.na(death), 0, death)) %>%
  dplyr::arrange(confirmed) %>%
  dplyr::ungroup() %>%
  dplyr::mutate(country = dplyr::if_else(country == "United Arab Emirates", "UAE", country)) %>%
  dplyr::mutate(country = dplyr::if_else(country == "Mainland China", "China", country)) %>%
  dplyr::mutate(country = dplyr::if_else(country == "North Macedonia", "N.Macedonia", country)) %>%
  dplyr::mutate(country = trimws(country)) %>%
  dplyr::mutate(country = factor(country, levels = country))
plotly::plot_ly(
  data = df_EU,
  x = ~country,
  # y = ~unrecovered,
  y = ~ confirmed,
  # text =  ~ confirmed,
  # textposition = 'auto',
  type = "bar",
  name = "Confirmed",
  marker = list(color = active_color)
) %>%
  plotly::add_trace(
    y = ~death,
    # text =  ~ death,
    # textposition = 'auto',
    name = "Dead",
    marker = list(color = death_color)
  ) %>%
  plotly::layout(
    barmode = "stack",
    yaxis = list(title = "Total cases"),
    xaxis = list(title = ""),
    hovermode = "compare",
    margin = list(
      # l = 60,
      # r = 40,
      b = 10,
      t = 10,
      pad = 2
    )
  )
```


Map
=======================================================================

### **World map of cases**

```{r}
# map tab
library(leaflet)
library(leafpop)
library(purrr)
cv_data_for_plot <- coronavirus %>%
  # dplyr::filter(country == "Denmark") %>%
  dplyr::filter(cases > 0) %>%
  dplyr::group_by(country, lat, long, type) %>%
  dplyr::summarise(cases = sum(cases)) %>%
  dplyr::mutate(log_cases = 2 * log(cases)) %>%
  dplyr::ungroup()
cv_data_for_plot.split <- cv_data_for_plot %>% split(cv_data_for_plot$type)
pal <- colorFactor(c("orange", "red", "green"), domain = c("confirmed", "death", "recovered"))
map_object <- leaflet() %>% addProviderTiles(providers$Stamen.Toner)
names(cv_data_for_plot.split) %>%
  purrr::walk(function(df) {
    map_object <<- map_object %>%
      addCircleMarkers(
        data = cv_data_for_plot.split[[df]],
        lng = ~long, lat = ~lat,
        #                 label=~as.character(cases),
        color = ~ pal(type),
        stroke = FALSE,
        fillOpacity = 0.8,
        radius = ~log_cases,
        popup = leafpop::popupTable(cv_data_for_plot.split[[df]],
          feature.id = FALSE,
          row.numbers = FALSE,
          zcol = c("type", "cases", "country")
        ),
        group = df,
        #                 clusterOptions = markerClusterOptions(removeOutsideVisibleBounds = F),
        labelOptions = labelOptions(
          noHide = F,
          direction = "auto"
        )
      )
  })
map_object %>%
  addLayersControl(
    overlayGroups = names(cv_data_for_plot.split),
    options = layersControlOptions(collapsed = FALSE)
  )
```


Forecasting Danish deaths
=======================================================================

Row {data-width=400}
-----------------------------------------------------------------------

### **ARIMA - AutoRegressive-Integrated-Moving-Average - for Danish deaths**
    
```{r forecast}
library(forecast)
library(tidyverse)

forecast_data <- df_daily# %>% slice(1:(n()-4))

min_data <- min(forecast_data$date)
max_data <- max(forecast_data$date)
forecast_length = 30

TS_data <- ts(forecast_data[,c(1,7)], frequency = 7)
TS_fit <- auto.arima(TS_data[,2])
TS_fc <- forecast(TS_fit, h=forecast_length)

fore.dates <- seq.Date(from = max_data+1,
                       to = max_data+forecast_length,
                       by="days")

plot_daily <- forecast_data[forecast_data$date >= "2020-02-27",]

TS_fc$mean <- round(TS_fc$mean, digits=0)
TS_fc$upper <- round(TS_fc$upper, digits=0)
TS_fc$lower <- round(TS_fc$lower, digits=0)

td <- plot_daily$death_cum[nrow(plot_daily)]

TS_fc$lower[TS_fc$lower%
    plotly::add_lines(x = plot_daily$date, y = plot_daily$death_cum,
                      color = I("black"),
                      name = "Observed deaths",
                      marker=list(mode='lines')) %>%
    plotly::add_lines(x = fore.dates, y = TS_fc$mean, color = I("blue"), name = "Predicted deaths") %>%
    plotly::add_ribbons(x = fore.dates,
                        ymin = TS_fc$lower[, 2],
                        ymax = TS_fc$upper[, 2],
                        color = I("gray95"),
                        name = "95% confidence") %>%
    plotly::add_ribbons(x = fore.dates,
                        ymin = TS_fc$lower[, 1],
                        ymax = TS_fc$upper[, 1],
                        color = I("gray80"), name = "80% confidence") %>% 
    layout(annotations = 
               list(x = 0.40, y = 0.90, text = paste("AIC:",TS_fit$aic, "| Log-Likelihood:", TS_fit$loglik, sep = " "), 
                    showarrow = F, xref='paper', yref='paper', 
                    xanchor='right', yanchor='auto', xshift=0, yshift=-14,
                    font=list(size=10, color="black"))
    )
```


Estimating R0
=======================================================================

Row {data-width=400}
-----------------------------------------------------------------------

### **Estimating Danish R0 using SIR model rougly equivalent to SSI / CDC estimates**
    
```{r}
library(tidyverse)
library(EpiEstim)
library(plotly)

url <- "https://api.covid19data.dk/ssi_newly_hospitalized"
hosp_raw <- jsonlite::fromJSON(url)
hosp_raw <- hosp_raw %>%
    as_tibble() %>%
    mutate(date = date %>% lubridate::ymd_hms() %>% as.Date())

ads <- hosp_raw %>%
      mutate(rollingmean = c(rep(NA, 3),  # Adding a rolling mean
                           zoo::rollmean(x = newly_hospitalized, k = 7), 
                           rep(NA, 3)))
    #slice(1:(n()-4)) %>% # remove last four entries


## Choosing parameters for estimating R ########################################

# We need to supply the estimate_R function with some parameters in order for it
# to estimate R. We will use two sets of parameters:
# 1) from the CDC where they have estimated mean and the standard deviation for
#    the serial interval.
# 2) Numbers I have learned that the danish SSI uses. I have only mean informed 
#    of their mean, so the standard deviation I have chosen to be the same as
#    the mean. This number can be changed as we get more info from SSI.

# CDC
# https://wwwnc.cdc.gov/eid/article/26/6/20-0357_article
cdc_mean_si = 3.96
cdc_std_si = 4.75

# SSI - kan arbejde med 5-6 (jf. WHO p. 2: https://www.who.int/docs/default-source/coronaviruse/situation-reports/20200306-sitrep-46-covid-19.pdf?sfvrsn=96b04adf_4)
ssi_mean_si = 4.7
ssi_std_si = 4.7

confirmed_cases <- ads  %>% drop_na() %>% select(I = rollingmean)

cdc_R <- estimate_R(confirmed_cases, 
                    method = "parametric_si", 
                    config = make_config(list(mean_si = cdc_mean_si, 
                                              std_si = cdc_std_si)))

ssi_R <- estimate_R(confirmed_cases, 
                    method = "parametric_si", 
                    config = make_config(list(mean_si = ssi_mean_si, 
                                              std_si = ssi_std_si)))

cdc_pd <- cdc_R$R %>% 
    as_tibble() %>% 
    select(t = t_end, 
           R0 = `Mean(R)`, 
           lower = `Quantile.0.05(R)`, 
           upper = `Quantile.0.95(R)`) %>%
    mutate(source = "cdc") 

ssi_pd <- ssi_R$R %>% 
    as_tibble() %>% 
    select(t = t_end, 
           R0 = `Mean(R)`, 
           lower = `Quantile.0.05(R)`, 
           upper = `Quantile.0.95(R)`) %>%
    mutate(source = "ssi") 

pd <- bind_rows(cdc_pd, ssi_pd) %>%
    mutate(t = (t-1) + as.Date("2020-02-28"))

pd$Date <- pd$t

facetlabs <- c(cdc = "Danish R0 using Center for Disease Control inputs",
               ssi = "Danish R0 using Statens Serum Institut inputs")

p1 <- ggplot(pd) + 
    geom_ribbon(aes(x = Date, ymin = lower, ymax = upper), fill = "lightgrey") + 
    geom_line(aes(x = Date, y = R0)) + 
    geom_hline(yintercept = 1, linetype = "dashed") + 
    facet_wrap(~ source, ncol = 1, 
               labeller = labeller(source = facetlabs)) + 
    theme_minimal() + 
    labs(x = "", y = "") + 
    scale_x_date(breaks = "weeks", labels = scales::date_format("%d-%m"))

ggplotly(p1)
  
```

Please note that the above graphs are adapted versions of estimates created by [Mikkel Krogsholm](https://gist.github.com/mikkelkrogsholm/df208ab854e3c13ab07d23c027af7b5b).

About
=======================================================================

**The Coronavirus Dashboard: the case of Denmark**

This Coronavirus dashboard: the case of Denmark provides an overview of the 2019 Novel Coronavirus COVID-19 (2019-nCoV) epidemic for Denmark. This dashboard is built with R using the R markdown-framework and was adapted from this [dashboard](https://ramikrispin.github.io/coronavirus_dashboard/){target="_blank"} by Rami Krispin.



The Dashboard has been further adapted per the dashboard developed by [Antoine Soetewey](https://github.com/AntoineSoetewey).



The part on estimating R-naught/R0 is based on code from [Mikkel Krogsholm](https://gist.github.com/mikkelkrogsholm/df208ab854e3c13ab07d23c027af7b5b).

**Data**


The input data for this dashboard is the dataset available from the [`{coronavirus}`](https://github.com/RamiKrispin/coronavirus){target="_blank"} R package. Make sure to download the development version of the package to have the latest data:
```
install.packages("devtools")
devtools::install_github("RamiKrispin/coronavirus")
```
The data and dashboard is refreshed on an almost daily basis.

The raw data is pulled from the Johns Hopkins University Center for Systems Science and Engineering (JHU CCSE) Coronavirus [repository](https://github.com/RamiKrispin/coronavirus-csv){target="_blank"}.

**Update**


The data is as of `r format(max(coronavirus$date))` and the dashboard has been updated on `r format(Sys.time())`.